2 * Author......: Jens Steube <jens.steube@gmail.com>
8 #include "include/constants.h"
9 #include "include/kernel_vendor.h"
24 #include "include/kernel_functions.c"
26 #include "common_nv.c"
29 #define VECT_COMPARE_S "check_single_vect1_comp4.c"
30 #define VECT_COMPARE_M "check_multi_vect1_comp4.c"
34 #define VECT_COMPARE_S "check_single_vect2_comp4.c"
35 #define VECT_COMPARE_M "check_multi_vect2_comp4.c"
39 #define VECT_COMPARE_S "check_single_vect4_comp4.c"
40 #define VECT_COMPARE_M "check_multi_vect4_comp4.c"
43 __device__ __constant__ comb_t c_combs[1024];
45 __device__ __constant__ u64 k_sha512[80] =
47 SHA512C00, SHA512C01, SHA512C02, SHA512C03,
48 SHA512C04, SHA512C05, SHA512C06, SHA512C07,
49 SHA512C08, SHA512C09, SHA512C0a, SHA512C0b,
50 SHA512C0c, SHA512C0d, SHA512C0e, SHA512C0f,
51 SHA512C10, SHA512C11, SHA512C12, SHA512C13,
52 SHA512C14, SHA512C15, SHA512C16, SHA512C17,
53 SHA512C18, SHA512C19, SHA512C1a, SHA512C1b,
54 SHA512C1c, SHA512C1d, SHA512C1e, SHA512C1f,
55 SHA512C20, SHA512C21, SHA512C22, SHA512C23,
56 SHA512C24, SHA512C25, SHA512C26, SHA512C27,
57 SHA512C28, SHA512C29, SHA512C2a, SHA512C2b,
58 SHA512C2c, SHA512C2d, SHA512C2e, SHA512C2f,
59 SHA512C30, SHA512C31, SHA512C32, SHA512C33,
60 SHA512C34, SHA512C35, SHA512C36, SHA512C37,
61 SHA512C38, SHA512C39, SHA512C3a, SHA512C3b,
62 SHA512C3c, SHA512C3d, SHA512C3e, SHA512C3f,
63 SHA512C40, SHA512C41, SHA512C42, SHA512C43,
64 SHA512C44, SHA512C45, SHA512C46, SHA512C47,
65 SHA512C48, SHA512C49, SHA512C4a, SHA512C4b,
66 SHA512C4c, SHA512C4d, SHA512C4e, SHA512C4f,
69 __device__ static void sha512_transform (const u64x w0[4], const u64x w1[4], const u64x w2[4], const u64x w3[4], u64x digest[8])
97 #define ROUND_EXPAND() \
99 w0_t = SHA512_EXPAND (we_t, w9_t, w1_t, w0_t); \
100 w1_t = SHA512_EXPAND (wf_t, wa_t, w2_t, w1_t); \
101 w2_t = SHA512_EXPAND (w0_t, wb_t, w3_t, w2_t); \
102 w3_t = SHA512_EXPAND (w1_t, wc_t, w4_t, w3_t); \
103 w4_t = SHA512_EXPAND (w2_t, wd_t, w5_t, w4_t); \
104 w5_t = SHA512_EXPAND (w3_t, we_t, w6_t, w5_t); \
105 w6_t = SHA512_EXPAND (w4_t, wf_t, w7_t, w6_t); \
106 w7_t = SHA512_EXPAND (w5_t, w0_t, w8_t, w7_t); \
107 w8_t = SHA512_EXPAND (w6_t, w1_t, w9_t, w8_t); \
108 w9_t = SHA512_EXPAND (w7_t, w2_t, wa_t, w9_t); \
109 wa_t = SHA512_EXPAND (w8_t, w3_t, wb_t, wa_t); \
110 wb_t = SHA512_EXPAND (w9_t, w4_t, wc_t, wb_t); \
111 wc_t = SHA512_EXPAND (wa_t, w5_t, wd_t, wc_t); \
112 wd_t = SHA512_EXPAND (wb_t, w6_t, we_t, wd_t); \
113 we_t = SHA512_EXPAND (wc_t, w7_t, wf_t, we_t); \
114 wf_t = SHA512_EXPAND (wd_t, w8_t, w0_t, wf_t); \
117 #define ROUND_STEP(i) \
119 SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha512[i + 0]); \
120 SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha512[i + 1]); \
121 SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha512[i + 2]); \
122 SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha512[i + 3]); \
123 SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha512[i + 4]); \
124 SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha512[i + 5]); \
125 SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha512[i + 6]); \
126 SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha512[i + 7]); \
127 SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha512[i + 8]); \
128 SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha512[i + 9]); \
129 SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha512[i + 10]); \
130 SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha512[i + 11]); \
131 SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha512[i + 12]); \
132 SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha512[i + 13]); \
133 SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, we_t, k_sha512[i + 14]); \
134 SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha512[i + 15]); \
139 for (int i = 16; i < 80; i += 16)
141 ROUND_EXPAND (); ROUND_STEP (i);
154 __device__ static void hmac_sha512_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u64x ipad[8], u64x opad[8])
161 w0_t[0] = hl32_to_64 (w0[0], w0[1]) ^ 0x3636363636363636;
162 w0_t[1] = hl32_to_64 (w0[2], w0[3]) ^ 0x3636363636363636;
163 w0_t[2] = hl32_to_64 (w1[0], w1[1]) ^ 0x3636363636363636;
164 w0_t[3] = hl32_to_64 (w1[2], w1[3]) ^ 0x3636363636363636;
165 w1_t[0] = hl32_to_64 (w2[0], w2[1]) ^ 0x3636363636363636;
166 w1_t[1] = hl32_to_64 (w2[2], w2[3]) ^ 0x3636363636363636;
167 w1_t[2] = hl32_to_64 (w3[0], w3[1]) ^ 0x3636363636363636;
168 w1_t[3] = hl32_to_64 (w3[2], w3[3]) ^ 0x3636363636363636;
169 w2_t[0] = 0 ^ 0x3636363636363636;
170 w2_t[1] = 0 ^ 0x3636363636363636;
171 w2_t[2] = 0 ^ 0x3636363636363636;
172 w2_t[3] = 0 ^ 0x3636363636363636;
173 w3_t[0] = 0 ^ 0x3636363636363636;
174 w3_t[1] = 0 ^ 0x3636363636363636;
175 w3_t[2] = 0 ^ 0x3636363636363636;
176 w3_t[3] = 0 ^ 0x3636363636363636;
187 sha512_transform (w0_t, w1_t, w2_t, w3_t, ipad);
189 w0_t[0] = hl32_to_64 (w0[0], w0[1]) ^ 0x5c5c5c5c5c5c5c5c;
190 w0_t[1] = hl32_to_64 (w0[2], w0[3]) ^ 0x5c5c5c5c5c5c5c5c;
191 w0_t[2] = hl32_to_64 (w1[0], w1[1]) ^ 0x5c5c5c5c5c5c5c5c;
192 w0_t[3] = hl32_to_64 (w1[2], w1[3]) ^ 0x5c5c5c5c5c5c5c5c;
193 w1_t[0] = hl32_to_64 (w2[0], w2[1]) ^ 0x5c5c5c5c5c5c5c5c;
194 w1_t[1] = hl32_to_64 (w2[2], w2[3]) ^ 0x5c5c5c5c5c5c5c5c;
195 w1_t[2] = hl32_to_64 (w3[0], w3[1]) ^ 0x5c5c5c5c5c5c5c5c;
196 w1_t[3] = hl32_to_64 (w3[2], w3[3]) ^ 0x5c5c5c5c5c5c5c5c;
197 w2_t[0] = 0 ^ 0x5c5c5c5c5c5c5c5c;
198 w2_t[1] = 0 ^ 0x5c5c5c5c5c5c5c5c;
199 w2_t[2] = 0 ^ 0x5c5c5c5c5c5c5c5c;
200 w2_t[3] = 0 ^ 0x5c5c5c5c5c5c5c5c;
201 w3_t[0] = 0 ^ 0x5c5c5c5c5c5c5c5c;
202 w3_t[1] = 0 ^ 0x5c5c5c5c5c5c5c5c;
203 w3_t[2] = 0 ^ 0x5c5c5c5c5c5c5c5c;
204 w3_t[3] = 0 ^ 0x5c5c5c5c5c5c5c5c;
215 sha512_transform (w0_t, w1_t, w2_t, w3_t, opad);
218 __device__ static void hmac_sha512_run (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u64x ipad[8], u64x opad[8], u64x digest[8])
225 w0_t[0] = hl32_to_64 (w0[0], w0[1]);
226 w0_t[1] = hl32_to_64 (w0[2], w0[3]);
227 w0_t[2] = hl32_to_64 (w1[0], w1[1]);
228 w0_t[3] = hl32_to_64 (w1[2], w1[3]);
229 w1_t[0] = hl32_to_64 (w2[0], w2[1]);
230 w1_t[1] = hl32_to_64 (w2[2], w2[3]);
231 w1_t[2] = hl32_to_64 (w3[0], w3[1]);
240 w3_t[3] = hl32_to_64 (w3[2], w3[3]);
251 sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
261 w2_t[0] = 0x8000000000000000;
268 w3_t[3] = (128 + 64) * 8;
279 sha512_transform (w0_t, w1_t, w2_t, w3_t, digest);
282 extern "C" __global__ void __launch_bounds__ (256, 1) m01750_m04 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
288 const u32 lid = threadIdx.x;
294 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
296 if (gid >= gid_max) return;
300 wordl0[0] = pws[gid].i[ 0];
301 wordl0[1] = pws[gid].i[ 1];
302 wordl0[2] = pws[gid].i[ 2];
303 wordl0[3] = pws[gid].i[ 3];
307 wordl1[0] = pws[gid].i[ 4];
308 wordl1[1] = pws[gid].i[ 5];
309 wordl1[2] = pws[gid].i[ 6];
310 wordl1[3] = pws[gid].i[ 7];
326 const u32 pw_l_len = pws[gid].pw_len;
328 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
330 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, c_combs[0].pw_len);
339 salt_buf0[0] = salt_bufs[salt_pos].salt_buf[ 0];
340 salt_buf0[1] = salt_bufs[salt_pos].salt_buf[ 1];
341 salt_buf0[2] = salt_bufs[salt_pos].salt_buf[ 2];
342 salt_buf0[3] = salt_bufs[salt_pos].salt_buf[ 3];
346 salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4];
347 salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5];
348 salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6];
349 salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7];
351 const u32 salt_len = salt_bufs[salt_pos].salt_len;
357 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
359 const u32 pw_r_len = c_combs[il_pos].pw_len;
361 const u32 pw_len = pw_l_len + pw_r_len;
365 wordr0[0] = c_combs[il_pos].i[0];
366 wordr0[1] = c_combs[il_pos].i[1];
367 wordr0[2] = c_combs[il_pos].i[2];
368 wordr0[3] = c_combs[il_pos].i[3];
372 wordr1[0] = c_combs[il_pos].i[4];
373 wordr1[1] = c_combs[il_pos].i[5];
374 wordr1[2] = c_combs[il_pos].i[6];
375 wordr1[3] = c_combs[il_pos].i[7];
391 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
393 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
398 w0[0] = wordl0[0] | wordr0[0];
399 w0[1] = wordl0[1] | wordr0[1];
400 w0[2] = wordl0[2] | wordr0[2];
401 w0[3] = wordl0[3] | wordr0[3];
405 w1[0] = wordl1[0] | wordr1[0];
406 w1[1] = wordl1[1] | wordr1[1];
407 w1[2] = wordl1[2] | wordr1[2];
408 w1[3] = wordl1[3] | wordr1[3];
412 w2[0] = wordl2[0] | wordr2[0];
413 w2[1] = wordl2[1] | wordr2[1];
414 w2[2] = wordl2[2] | wordr2[2];
415 w2[3] = wordl2[3] | wordr2[3];
419 w3[0] = wordl3[0] | wordr3[0];
420 w3[1] = wordl3[1] | wordr3[1];
421 w3[2] = wordl3[2] | wordr3[2];
422 w3[3] = wordl3[3] | wordr3[3];
430 w0_t[0] = swap_workaround (w0[0]);
431 w0_t[1] = swap_workaround (w0[1]);
432 w0_t[2] = swap_workaround (w0[2]);
433 w0_t[3] = swap_workaround (w0[3]);
437 w1_t[0] = swap_workaround (w1[0]);
438 w1_t[1] = swap_workaround (w1[1]);
439 w1_t[2] = swap_workaround (w1[2]);
440 w1_t[3] = swap_workaround (w1[3]);
459 hmac_sha512_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
461 w0_t[0] = swap_workaround (salt_buf0[0]);
462 w0_t[1] = swap_workaround (salt_buf0[1]);
463 w0_t[2] = swap_workaround (salt_buf0[2]);
464 w0_t[3] = swap_workaround (salt_buf0[3]);
465 w1_t[0] = swap_workaround (salt_buf1[0]);
466 w1_t[1] = swap_workaround (salt_buf1[1]);
467 w1_t[2] = swap_workaround (salt_buf1[2]);
468 w1_t[3] = swap_workaround (salt_buf1[3]);
476 w3_t[3] = (128 + salt_len) * 8;
480 hmac_sha512_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
483 const u32x r0 = l32_from_64 (digest[7]);
484 const u32x r1 = h32_from_64 (digest[7]);
485 const u32x r2 = l32_from_64 (digest[3]);
486 const u32x r3 = h32_from_64 (digest[3]);
488 #include VECT_COMPARE_M
492 extern "C" __global__ void __launch_bounds__ (256, 1) m01750_m08 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
496 extern "C" __global__ void __launch_bounds__ (256, 1) m01750_m16 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
500 extern "C" __global__ void __launch_bounds__ (256, 1) m01750_s04 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
506 const u32 lid = threadIdx.x;
512 const u32 gid = (blockIdx.x * blockDim.x) + threadIdx.x;
514 if (gid >= gid_max) return;
518 wordl0[0] = pws[gid].i[ 0];
519 wordl0[1] = pws[gid].i[ 1];
520 wordl0[2] = pws[gid].i[ 2];
521 wordl0[3] = pws[gid].i[ 3];
525 wordl1[0] = pws[gid].i[ 4];
526 wordl1[1] = pws[gid].i[ 5];
527 wordl1[2] = pws[gid].i[ 6];
528 wordl1[3] = pws[gid].i[ 7];
544 const u32 pw_l_len = pws[gid].pw_len;
546 if (combs_mode == COMBINATOR_MODE_BASE_RIGHT)
548 switch_buffer_by_offset (wordl0, wordl1, wordl2, wordl3, c_combs[0].pw_len);
557 salt_buf0[0] = salt_bufs[salt_pos].salt_buf[ 0];
558 salt_buf0[1] = salt_bufs[salt_pos].salt_buf[ 1];
559 salt_buf0[2] = salt_bufs[salt_pos].salt_buf[ 2];
560 salt_buf0[3] = salt_bufs[salt_pos].salt_buf[ 3];
564 salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4];
565 salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5];
566 salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6];
567 salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7];
569 const u32 salt_len = salt_bufs[salt_pos].salt_len;
575 const u32 search[4] =
577 digests_buf[digests_offset].digest_buf[DGST_R0],
578 digests_buf[digests_offset].digest_buf[DGST_R1],
579 digests_buf[digests_offset].digest_buf[DGST_R2],
580 digests_buf[digests_offset].digest_buf[DGST_R3]
587 for (u32 il_pos = 0; il_pos < combs_cnt; il_pos++)
589 const u32 pw_r_len = c_combs[il_pos].pw_len;
591 const u32 pw_len = pw_l_len + pw_r_len;
595 wordr0[0] = c_combs[il_pos].i[0];
596 wordr0[1] = c_combs[il_pos].i[1];
597 wordr0[2] = c_combs[il_pos].i[2];
598 wordr0[3] = c_combs[il_pos].i[3];
602 wordr1[0] = c_combs[il_pos].i[4];
603 wordr1[1] = c_combs[il_pos].i[5];
604 wordr1[2] = c_combs[il_pos].i[6];
605 wordr1[3] = c_combs[il_pos].i[7];
621 if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
623 switch_buffer_by_offset (wordr0, wordr1, wordr2, wordr3, pw_l_len);
628 w0[0] = wordl0[0] | wordr0[0];
629 w0[1] = wordl0[1] | wordr0[1];
630 w0[2] = wordl0[2] | wordr0[2];
631 w0[3] = wordl0[3] | wordr0[3];
635 w1[0] = wordl1[0] | wordr1[0];
636 w1[1] = wordl1[1] | wordr1[1];
637 w1[2] = wordl1[2] | wordr1[2];
638 w1[3] = wordl1[3] | wordr1[3];
642 w2[0] = wordl2[0] | wordr2[0];
643 w2[1] = wordl2[1] | wordr2[1];
644 w2[2] = wordl2[2] | wordr2[2];
645 w2[3] = wordl2[3] | wordr2[3];
649 w3[0] = wordl3[0] | wordr3[0];
650 w3[1] = wordl3[1] | wordr3[1];
651 w3[2] = wordl3[2] | wordr3[2];
652 w3[3] = wordl3[3] | wordr3[3];
660 w0_t[0] = swap_workaround (w0[0]);
661 w0_t[1] = swap_workaround (w0[1]);
662 w0_t[2] = swap_workaround (w0[2]);
663 w0_t[3] = swap_workaround (w0[3]);
667 w1_t[0] = swap_workaround (w1[0]);
668 w1_t[1] = swap_workaround (w1[1]);
669 w1_t[2] = swap_workaround (w1[2]);
670 w1_t[3] = swap_workaround (w1[3]);
689 hmac_sha512_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad);
691 w0_t[0] = swap_workaround (salt_buf0[0]);
692 w0_t[1] = swap_workaround (salt_buf0[1]);
693 w0_t[2] = swap_workaround (salt_buf0[2]);
694 w0_t[3] = swap_workaround (salt_buf0[3]);
695 w1_t[0] = swap_workaround (salt_buf1[0]);
696 w1_t[1] = swap_workaround (salt_buf1[1]);
697 w1_t[2] = swap_workaround (salt_buf1[2]);
698 w1_t[3] = swap_workaround (salt_buf1[3]);
706 w3_t[3] = (128 + salt_len) * 8;
710 hmac_sha512_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest);
713 const u32x r0 = l32_from_64 (digest[7]);
714 const u32x r1 = h32_from_64 (digest[7]);
715 const u32x r2 = l32_from_64 (digest[3]);
716 const u32x r3 = h32_from_64 (digest[3]);
718 #include VECT_COMPARE_S
722 extern "C" __global__ void __launch_bounds__ (256, 1) m01750_s08 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
726 extern "C" __global__ void __launch_bounds__ (256, 1) m01750_s16 (const pw_t *pws, const gpu_rule_t *rules_buf, const comb_t *combs_buf, const bf_t *bfs_buf, const void *tmps, void *hooks, const u32 *bitmaps_buf_s1_a, const u32 *bitmaps_buf_s1_b, const u32 *bitmaps_buf_s1_c, const u32 *bitmaps_buf_s1_d, const u32 *bitmaps_buf_s2_a, const u32 *bitmaps_buf_s2_b, const u32 *bitmaps_buf_s2_c, const u32 *bitmaps_buf_s2_d, plain_t *plains_buf, const digest_t *digests_buf, u32 *hashes_shown, const salt_t *salt_bufs, const void *esalt_bufs, u32 *d_return_buf, u32 *d_scryptV_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 combs_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)